Real-time Tracking Algorithm of Vehicle in Dense Clutter

Yang Le, Zhang Chang-yong, Ping XuResearch 2015

Abstract

A speedy algorithm was proposed to aim at the problem of vehicle tracking in dense clutter. Firstly,illumination compensation was used on video to reduce illumination effects. The moving target was detected with Surendra algorithm and three-frame-difference method. The ROI was established to narrow the identified range of Haar classifier and used to improve efficiency of the algorithm. Secondly,the moving vehicle was identified in ROI by the cascade classifier with the high detection rate but low false detection rate. Thirdly,the moving vehicle was tracked with the Camshift algorithm. An embedded system platform was established to verify the effect of algorithm. Experimental results showed that the algorithm had good reliability and robustness,high detection rate and low false detection rate compared with single traditional algorithm.
BibTeX
@inproceedings{le2015real,
  author = {Yang Le and Zhang Chang-yong and Ping Xu},
  title = {{Real-time Tracking Algorithm of Vehicle in Dense Clutter}},
  booktitle = {Research},
  year = {2015},
}

Loading related content...

Figure